420 BOTANICAL GAZETTE [JUNE 
reported cases of apogamous polyembryony from the nucellar 
tissue just above the embryo sac. In fig. 32, the most mature 
embryo represented, a distinct development of tissue in the 
suspensor region is shown, which appears late in the develop- 
ment of the embryo, and is well marked off from the embryo 
proper by a narrow neck. This suspensor tissue is erythrophilous 
as compared with the embryo, showing its closer relation to 
nutritive supplies. From this late development of tissue in the 
suspensor region, and its great activity, it would suggest its 
possible association with supposed cases of polyembryony. 
DEVELOPMENT OF ENDOSPERM. 
It is becoming well known that the first division of the 
definitive nucleus holds no direct relation to the fusion of the 
sexual nuclei. It may precede or follow this fusion, or be 
coincident with it. So far as observed it occurs after the 
entrance of the pollen tube into the sac, but at almost any time 
thereafter, apparently related to no special one of those events 
which follow and which go to make up the process of fertilization. 
It seems reasonable to suppose that the inciting event is the 
entrance of the tube. In the case of Lilium the sexual and polar 
pairs of nuclei were observed to fuse simultaneously, but when 
division begins the endosperm nuclei divide more rapidly than 
do the cells of the embryo. When the embryo is but two or 
three-celled numerous free endosperm nuclei are scattered 
throughout the embryo sac (fig. 35). Later cell walls begin to 
form in the endosperm. A very interesting phase in the division 
of the definitive nucleus is shown in fig. 34. The remarkably 
distinct radiations about the forming daughter nuclei seem to be 
due to the spindle fibers pulled apart, and to other radiations 
which are similar to those which appear about the forming 
da : 
a hter nuclei in the divisions of the nuclei of the embryo sac 
ag which precede fertilization. Such radiations are difficult to inter- 
ee Lees but their distinctness j in this preparation could not be a 
Du ing the developn pment on 
- of the = embryo a ned endosperm ‘the ae 
